How Long Baidu Indexing Takes, and What Actually Speeds It Up

You submitted your domain to Baidu. A week later, nothing. Two weeks later, still nothing, and you are starting to wonder whether the submission went through at all.

That is the most common question we get about Baidu, and the answer is usually that nothing has gone wrong yet. Baidu is slow with a new foreign site, slower than Google by a wide margin. The useful question is not what the average wait is. It is which of the things you control are adding weeks to it.

Three things that get confused with each other

Before any timeline makes sense, separate them:

Discovery: Baidu knows the URL exists. Submission does this, and it is instant.
Crawling: Baiduspider actually fetches the page. This can fail silently for months.
Inclusion (收录): the page is in the index and can be returned in results. This is what a site: query shows.
Ranking: the page competes for a query. Different problem, mostly a content problem.

A page can be discovered on day one, crawled on day nine, and never included. Submitting more URLs does not fix that. Most people asking "how long does Baidu take" are actually stuck between crawling and inclusion.

The realistic timeline

For a typical European site: hosted in Europe, no ICP filing, content in English or a European language, no links from Chinese sites.

First Baiduspider visit: one day to three weeks after submission. Usually the homepage, and often nothing else for a while.
First page visible in a site: query: one to six weeks.
Useful coverage of a small site: one to three months.
Coverage that holds and does not shrink at the next re-evaluation: three to six months, and realistically only with Simplified Chinese pages on the site.

Compare that with a mainland-hosted site that has an ICP filing and Chinese content: homepage in the index within days, the rest within a couple of weeks. The gap between those two lines is what the rest of this article is about.

Baidu publishes no service level for any of this, and there is no support queue to escalate to. What you get is a set of channels with different throughput, and a crawl budget that grows or shrinks depending on how your server behaves.

First, check whether you are indexed at all

Do this before changing anything.

Go to baidu.com and search site:yourdomain.com. The result count is an estimate and it moves around, so read it as a yes or no rather than a measurement. If you get results, you are indexed and the question becomes coverage. If you get zero after six weeks, something is broken and waiting longer will not fix it.

Two more checks worth doing:

Search your exact homepage title in quotation marks. If the site is indexed but the site: operator is behaving oddly, this often surfaces it.
In Baidu Webmaster Tools (百度搜索资源平台) look at 索引量 (index volume) and 抓取频次 (crawl frequency). Those are the honest numbers. 抓取异常 (crawl errors) tells you whether the spider is getting timeouts, 403s or DNS failures, which is where most foreign sites find their actual problem.

If crawl frequency is near zero and crawl errors are not, you have a technical fault, not a patience problem.

The submission channels, and what each is actually worth

Baidu offers several ways to hand it a URL, and they are not equivalent.

Sitemap

Discovery only, and the slowest queue of the lot. Baidu treats a sitemap as a list of suggestions to work through when it gets round to it. Submit one, keep it accurate, do not expect it to move anything this week. It matters most for large sites where you need Baidu to find deep pages at all.

Manual URL submission (普通收录)

Paste URLs into the dashboard. Capped per day, per site. It is a hint, not an instruction. Useful for a handful of pages you genuinely care about, useless as a strategy.

Active push via API (主动推送)

The best of the ordinary channels. Your server posts new and updated URLs to Baidu's endpoint the moment they change, which cuts discovery latency to nothing and, in our experience, shortens the wait to a first crawl noticeably. The daily quota scales with how much of your site is already indexed, which is a chicken and egg problem at the start but stops being one quickly. If you publish regularly, wire this into your build or CMS and forget about it.

Automatic push (自动推送)

A JavaScript snippet that pings Baidu when a page is viewed. It needs real visitors from China to do anything, which a newly submitted foreign site does not have. It also adds a third-party script to every page, on a site where load time from inside China is already your weak point. Skip it until you have Chinese traffic.

Fast inclusion (快速收录)

This is the channel everyone means when they ask how to get indexed quickly, and the honest answer is uncomfortable.

快速收录 is a separate, quota-limited channel: a fixed number of URLs per day, and URLs pushed through it are processed far faster than anything in the ordinary queue. The quota is granted by Baidu, per site, based on its own assessment of the site's quality and mobile experience. It has moved between programme names over the years, and the current shape of it is a per-site allowance rather than something you switch on.

Three things follow from that. It is not for sale. It is not granted on request. And a brand new foreign-hosted site does not have it, because there is no crawl history for Baidu to judge. If it becomes available for your domain at all, it will be after months of clean crawls and Chinese-language content that Baidu keeps rather than drops. Anyone offering to sell you fast inclusion quota is selling you something other than what you think.

The practical order, then: active push for speed, sitemap for coverage, manual submission for the few pages that matter most, and fast inclusion only if your site has earned it.

Why hosting outside mainland China is slower

Baiduspider crawls from mainland IP ranges. Every request to your German or Irish server crosses the border, and that has three effects that compound.

Latency and loss. Round trips of 200 to 400 milliseconds are normal, and during peak hours packet loss on that path is real. A page that loads in 800 milliseconds from Frankfurt can take five seconds or more from Beijing.

Crawl budget shrinks. Baidu, like any crawler, backs off from slow or unreliable hosts. Timeouts and 5xx responses reduce your crawl frequency, which means fewer pages fetched per visit, which means slower inclusion. This is the mechanism behind most "Baidu just stopped crawling us" reports.

Blocked resources hang the page. Google Fonts, googleapis.com, Google Analytics, YouTube embeds, reCAPTCHA, Facebook widgets: all unreachable from inside China. A browser waits on each of them until it gives up. Anything Baidu fetches with rendering hits the same wall. Self-hosting fonts and scripts is the single cheapest fix available to a European company, and it usually takes an afternoon.

Short of a mainland presence, hosting or fronting the site from Hong Kong, Singapore or Japan cuts the round trip substantially. It is not the same as being inside, but it is a large improvement over Europe and it requires no Chinese entity.

What ICP filing changes, and what it does not

ICP备案 (ICP filing) is the record every site hosted on mainland servers must have with the Ministry of Industry and Information Technology. ICP许可证 (the ICP licence) is the heavier permission required for commercial online services. Both need a mainland Chinese legal entity, and filing also needs a mainland hosting provider, so a foreign company with no Chinese subsidiary and no local partner cannot obtain either.

Here is the part that gets misreported: Baidu does not require an ICP filing to index your site. A .com hosted in Frankfurt can be indexed and does get indexed, and no published rule treats the filing number as a ranking factor.

What filing actually unlocks is mainland hosting, and mainland hosting is what delivers the speed and the crawl budget. The advantage is real but indirect. If you have no Chinese entity, that is a reason to fix everything else on this list, not a reason to give up on being indexed.

Page speed, measured from the right place

The number that matters is how your page performs on a mobile connection inside China, not what your European monitoring dashboard says. Baidu has been explicit for years that mobile page experience feeds into how it treats a site, and the working target for a landing page is under three seconds to meaningful content on a mainland mobile network.

What a European team can act on today:

Render on the server. Baiduspider handles JavaScript far less reliably than Googlebot. If your text only exists after hydration, assume Baidu does not see it. Server-side rendering or static HTML is not optional here.
Self-host fonts, icons and analytics. Every request to a blocked host is a stall.
Compress images properly and set long cache headers on static assets.
Keep redirects to one hop. Each chain link costs another cross-border round trip.
Serve one URL per page. No session parameters, no tracking parameters in internal links, one canonical form of the domain.

Measure it from there, not from here: several mainland testing services will fetch your URL from a dozen Chinese cities and give you real numbers.

Is Baiduspider even allowed in?

Open yourdomain.com/robots.txt right now. We have lost count of the sites that submitted to Baidu, waited two months and were blocking the crawler the whole time.

The usual causes:

A robots.txt from a template that names Googlebot and Bingbot as allowed and then ends with User-agent: * and Disallow: /.
An explicit User-agent: Baiduspider / Disallow: / block that somebody added years ago because Baiduspider was crawling too aggressively. It is very common on older European sites.
WAF or CDN bot protection, which is the most frequent cause and the one nobody checks, because it does not appear in robots.txt at all. Managed bot rules challenge or 403 unfamiliar user agents, and some setups geo-block mainland Chinese IP ranges outright. As far as Baidu is concerned, your site returns an error.
A noindex meta tag or X-Robots-Tag header left over from staging.
A rel="canonical" pointing at a different domain.

Baiduspider identifies itself as Baiduspider with desktop and mobile variants, plus Baiduspider-image, Baiduspider-video and Baiduspider-news. A genuine one resolves by reverse DNS under baidu.com or baidu.jp. Anything claiming to be Baiduspider from another host is a spoof, and blocking those costs you nothing.

A quick check you can run in a terminal:

curl -A "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)" -I https://yourdomain.com/

A 200 is what you want. A 403, a redirect to a challenge page or a CAPTCHA means the crawler is being turned away. That test does not cover geo-blocking, though, so also ask whoever runs your CDN what the rules do with traffic from mainland China.

Why pages get indexed and then vanish

Index volume goes down as well as up, and a drop is not a bug. Baidu re-evaluates and quietly drops pages it decides are not worth keeping: English-only pages with no Chinese audience, thin pages, near-duplicates across language versions, auto-translated text.

This is where a foreign site's real ceiling sits. Every technical fix in this article buys you crawl speed. What decides whether inclusion holds is whether there is Simplified Chinese content on the site that a Chinese reader would actually want. Machine translation does not clear that bar, and Baidu is noticeably better at spotting it than it used to be.

What you control and what you do not

You control: robots.txt, your WAF and CDN rules, third-party resources, server-side rendering, load time from China, hosting location, sitemap accuracy, active push, one canonical URL per page, and whether Chinese content exists at all.

You do not control: Baidu's crawl schedule, how fast the queue moves, whether fast inclusion quota is ever granted, or what index volume settles at. None of it can be bought, and there is no support ticket that changes it.

A schedule that is not wishful thinking

Week 0: submit, verify the domain, submit the sitemap, wire up active push, and check robots.txt and CDN rules before anything else.
Week 1: look at crawl frequency and crawl errors. Fix whatever the error report shows. Do not touch the content yet.
Weeks 2 to 3: first site: check. Homepage only is normal and fine.
Week 6: if site: still returns nothing and the crawl log shows no successful fetches, stop waiting and treat it as a technical fault. It almost always is.
Weeks 8 to 12: track index volume rather than the site: count. Start adding Simplified Chinese pages.
Every quarter after that: re-check crawl errors, re-check that nothing new is blocking the spider, and see whether coverage is growing or shrinking.

If you would rather hand it over

Most of the wait is not mysterious. It is a blocked crawler, a slow page from Beijing, or a site with nothing on it that a Chinese reader wants, and those are fixable.

The submission itself is the part that stops most European companies, because a Baidu account requires a mainland Chinese mobile number to register. Our Baidu submission service takes your domain, runs the whole submission and reports back on what Baidu actually indexed, at a fixed price per domain. You need no Chinese phone number, no Chinese entity and no Baidu account of your own. If you also want the technical side fixed before you submit, book a call and we will scope it.